Upload app.py

app.py

@@ -12,13 +12,16 @@ from transformers import (
 import streamlit as st
 import openai
 
+
 @st.experimental_singleton
 def get_data():
     data = pd.read_csv("earnings_calls_sentencewise.csv")
     return data
 
+
 # Initialize models from HuggingFace
 
+
 @st.experimental_singleton
 def get_t5_model():
     return pipeline("summarization", model="t5-small", tokenizer="t5-small")
@@ -69,6 +72,7 @@ def format_query(query_results):
     context = [result["metadata"]["Text"] for result in query_results["matches"]]
     return context
 
+
 def sentence_id_combine(data, query_results, lag=2):
     # Extract sentence IDs from query results
     ids = [result["metadata"]["Sentence_id"] for result in query_results["matches"]]
@@ -82,14 +86,16 @@ def sentence_id_combine(data, query_results, lag=2):
     ]
     # Create a list of context sentences by joining the sentences corresponding to the lookup IDs
     context_list = [
-        " ".join(data.Text.iloc[lookup_id].to_list()) for lookup_id in lookup_ids
+        ". ".join(data.Text.iloc[lookup_id].to_list()) for lookup_id in lookup_ids
     ]
     return context_list
 
+
 def text_lookup(data, sentence_ids):
-    context = " ".join(data.iloc[sentence_ids].to_list())
+    context = ". ".join(data.iloc[sentence_ids].to_list())
     return context
 
+
 def gpt3_summary(text):
     response = openai.Completion.create(
         model="text-davinci-003",
@@ -163,7 +169,7 @@ window = int(st.number_input("Sentence Window Size", 1, 3, value=1))
 
 data = get_data()
 
-#context_list = format_query(query_results)
+# context_list = format_query(query_results)
 context_list = sentence_id_combine(data, query_results, lag=window)
 
 
@@ -181,10 +187,10 @@ if decoder_model == "GPT3 (text_davinci)":
     output_text = []
     for context_text in context_list:
         output_text.append(gpt3_summary(context_text))
-    generated_text = " ".join(output_text)
+    generated_text = ". ".join(output_text)
     st.write(gpt3_summary(generated_text))
 
-elif decoder_model == "GPT3 - QA":
+elif decoder_model == "GPT3 (QA_davinci)":
     openai_key = st.text_input(
         "Enter OpenAI key",
         value="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",
@@ -195,7 +201,7 @@ elif decoder_model == "GPT3 - QA":
     output_text = []
     for context_text in context_list:
         output_text.append(gpt3_qa(query_text, context_text))
-    generated_text = " ".join(output_text)
+    generated_text = ". ".join(output_text)
     st.write(gpt3_qa(query_text, generated_text))
 
 elif decoder_model == "T5":
@@ -203,7 +209,7 @@ elif decoder_model == "T5":
     output_text = []
     for context_text in context_list:
         output_text.append(t5_pipeline(context_text)[0]["summary_text"])
-    generated_text = " ".join(output_text)
+    generated_text = ". ".join(output_text)
     st.write(t5_pipeline(generated_text)[0]["summary_text"])
 
 elif decoder_model == "FLAN-T5":
@@ -211,7 +217,7 @@ elif decoder_model == "FLAN-T5":
     output_text = []
     for context_text in context_list:
         output_text.append(flan_t5_pipeline(context_text)[0]["summary_text"])
-    generated_text = " ".join(output_text)
+    generated_text = ". ".join(output_text)
     st.write(flan_t5_pipeline(generated_text)[0]["summary_text"])
 
 show_retrieved_text = st.checkbox("Show Retrieved Text", value=False)